On Tue, Oct 27, 2009 at 11:54 PM, <j...@google.com> wrote: > > http://gwt-code-reviews.appspot.com/86809/diff/1/2#newcode119 > Line 119: <echo message=" or -Dgwt.selenium.hosts, but that's all."/> > On 2009/10/28 03:00:56, zundel wrote: > >> I know this is just temporary, but its too bad the top level test >> > target doesn't > >> actually fail if no tests run. Its hard to believe a human being >> > As John notes, it's not true that "no tests ran," and it is true that the tests we specify did run... it's just that that set is unexpected.
> Frankly, I hardly look at the ant output unless the build fails. > > That was my point against John's thought that noticing "few" tests ran was enough, yes. > Testing each of the properties named in the echo message with a >> > <condition> is burdensome, I know, but maybe worth it because this is such > a > deviation from what we're used to. > > It seems bad to always print this message even if the user supplied the > hosts for RBM/Selenium. > For both of you on this point, I'm not at all worried about "burdensome"... but I am worried about expectation. It "should be" that if you set the properties and run test (vice running test.remote and/or test.selenium explicitly), you test remote AND htmlunit. Since it won't do that now, do we still warn if you the props? If so, is it much different than not bothering with the conditions? --~--~---------~--~----~------------~-------~--~----~ http://groups.google.com/group/Google-Web-Toolkit-Contributors -~----------~----~----~----~------~----~------~--~---
